Considerations for implementing OKRs in your organization

Things to know before you embark on your OKRs adventure.

So, you’re ready: you’re convinced of the benefits that OKRs can bring to your organization, you’ve secured the support of your CEO or the head of your department, and you’ve got a good idea of where you want to go with this. Now what? How do you get started? And what are the dangers you need to be aware of?

In this series, we are going to look at three key aspects of implementing OKRs in your organization:

  1. Things to know before you start
  2. How to build good OKRs, and
  3. Why tracking OKRs is as important as building them (and how to run an OKR cycle)

So let's start with the first topic: Things to know before you start implementing OKRs in your organization.

It's harder than you think

One of the great features of the OKR approach is how simple the idea fundamentally is (you can explain it in under 3 minutes), and how it adapts to the peculiarities of each company. But this apparent simplicity can blind you to the fact that, for all its apparent simplicity, goal setting through OKRs is quite a subtle approach. To reap the benefits, the organization needs to change its mindset about goals and objectives, and learn how to tackle the practical challenges of implementing OKRs.

As Christina Wodtke explains in her book Radical Focus, “When I work with clients to implement OKRs, I give them a warning: you will fail the first time. They do all fail, but they all fail in their own special way. [...] However, the successful ones all have the same characteristic: they try again. The only hope for success is iteration. This does not mean blindly trying the same thing over and over again. I believe that is the definition of insanity. Instead, you track closely what works, and what does not, and you do more of what works and less of what doesn’t. The heart of success is learning.

So what kind of pitfalls should you expect to come across (or fall into…)?

The main source of issues lies in not giving enough thought to your company objectives: most companies have pretty clear “classical” company objectives, and it’s easy to just “translate” them in to OKRs and be done with it. But the shape of your company OKRs is going to have a massive impact on the definition of most (if not all) OKRs at lower levels. If they don’t meet the criteria for “good” objectives, then you could find yourself trying to fit a square peg in a round hole over and over again as you cascade the OKRs down the organization.

As Mike Lee, cofounder and CEO of MyFitnessPal explains: “When we began to implement OKRs, it was harder than anticipated. We didn’t appreciate how much thought it took to create the right company objectives, and then to cascade them down to drive contributors’ behavior. We found it challenging to strike a balance between high-level, strategic thinking and more granular, directive communication.”

Issues will include:

  • Being waaaaaay to ambitious with your company objectives
  • Using overly vague language (or full of buzzwords that don’t have much meaning)
  • Or at the other end of the spectrum, setting “cut and dry” objectives that impose and direct, instead of being inspirational
  • Setting too many objectives (if you’ve got more than 3, you should be asking yourself questions)
  • Not involving the whole of the upper management in designing company OKRs, which will often leave out key sections of the organization.

Another big issue that I’ve seen is upper management handing over a copy of one book or another to department heads and asking them to “go do their OKRs”. One of the key features of OKRs, one that brings (in my opinion) some of the biggest benefits, is the conversations that they generate inside the organization, top-down, bottom-up, across functions and geographies, sometimes in totally unexpected ways. If you don’t create forums where these conversations can occur, then you’re losing out on a big chunk of what OKRs can give you.

As a side note, if you don’t accompany and guide all the organization in learning to build and use OKRs, you run into the possibility that different people will understand completely different things. Which in turn means that any conversations could potentially end up as a dialogue of the deaf, with people thinking that the other has understood what they meant when the latter did not have the same frame of reference. The concept of OKRs is simple, but it still take *some* fine tuning until the machine works smoothly.

You can’t do it without executive commitment

If you follow the PlaniswareHub blog (or the PlaniswareHub Monthly Brief), then you know how vital executive commitment is to any initiative that requires change.

As Jini Kim, Cofounder and CEO of Nuna explains, “With hindsight I would have started [to implement OKRs] with our leadership team of five. For structured goal setting to prosper, as our company learned the hard way, executives need to commit to the process. It may take a quarter or two to overcome your managers’ resistance and get them acclimated to OKRs—to view them not as a necessary evil, or some perfunctory exercise, but as a practical tool to fulfill your organization’s top priorities.

Until your executives are fully on board, you can’t expect contributors to follow suit—especially when a company’s OKRs are aspirational. The more challenging an objective, the more tempting it may be to abandon it. People naturally look to their bosses in setting goals and following through. If the officers jump ship in the middle of a stormy voyage, you can’t expect the sailors to bring it into port.

That is why one of the top best practices for implementing OKRs is to use a phased roll-out starting with upper management first.

Upper management here will be the highest level of management involved in the OKR initiative. If OKRs are deployed in the whole company, then upper management will be the C-suite. But if OKRs are deployed only in a business unit or a department, then it will be whoever leads those entities.

This phase may take up anywhere between 6 months to a year.  Yes, that long.

Part of the reason is that, until you’ve actually started to have these conversations about the company’s objectives, things often look quite simple: everyone has a good idea of what the company’s three top objectives should be, and they expect the whole process to go smoothly and quickly. That is, until everyone writes down what those three objectives are. And then, mayhem ensues… And that’s because nobody has the same three objectives. Companies routinely report ending up with a list of anywhere between 15 to 85 *different* objectives.

But until upper management have had hands-on experience of wrestling with OKRs, learning to build and use them in their routine life, it will be difficult for them to appreciate what the exercise is about.

Just as importantly, having leaders work with OKRs first (and debugging the first issues) will strengthen their commitment (“It’s worth it! We’ve done it and it works!”) and allow the process to gain momentum.

“To inspire true commitment, leaders must practice what they teach. They must model the behavior they expect of others.” — Jini Kim, Cofounder and CEO of Nuna

The OKR Shepherd…

… or the art of strategically nipping at your colleagues’ heels when they fall behind on their OKRs.

As John Doerr explains, “For an OKR system to function effectively, the team deploying it— whether a group of top executives or an entire organization—must adopt it universally. No exceptions, no opt-outs. Yes, there will be late adopters, resisters, and garden-variety procrastinators. To prod them to join the flock, a best practice is to designate one or more OKR shepherds.”

I have to admit, I love the image of an OKR shepherd (a throwback to the pastoral ambitions of my ancestors, maybe?), nipping at the heels of the laggards and prodding them back to join the flock.

Here is an extract from a classic Google Ventures Startup Lab video where Rick Klau gives a perfect example of what such a nipping might look like in real life:

OKRs must be separate from individual compensation

Fundamentally, OKRs aim to build a culture of accountability, based on (among other ideas) radical transparency. This means that, for this to work, people should not be afraid to openly admit their missteps, otherwise they will be reluctant to take healthy risks. One part of that can be achieved through the modelling of upper management: if *they* are not afraid of admitting to their mistakes or unsuccess’s, then this opens the door for everyone to be a bit more candid about those moments when things didn’t quite work out.

But that cannot be all, because the accounting of successes and failures are often a big part in how compensation (raises and bonuses) are awarded. As I often say: “You can’t on the one hand ask people to fail 30% of the time, and then hold that against them when the time comes round to discuss bonuses.”

John Doerr is categorical about that one: “Divorce compensation (both raises and bonuses) from OKRs. These should be two distinct conversations, with their own cadences and calendars. The first is a backward-looking assessment, typically held at year’s end. The second is an ongoing, forward-looking dialogue between leaders and contributors. It centers on five questions: What are you working on? How are you doing; how are your OKRs coming along? Is there anything impeding your work? What do you need from me to be (more) successful? How do you need to grow to achieve your career goals?


At Google, according to Laszlo Bock, OKRs amount to a third or less of performance ratings. They take a backseat to feedback from cross-functional teams, and most of all to context. “It’s always possible—even with a goal-setting system—to get the goals wrong,” Laszlo says. “Maybe the market does something crazy, or a client leaves their job and suddenly you have to rebuild from scratch. You try to keep all of that in consideration.” Google is careful to segregate raw goal scores from compensation decisions. Their OKR numbers are actually wiped from the system after each cycle!”

To learn more about how Google articulates OKRs with the elements guiding compensation, read this article: Laszlo Bock: Divorce Compensation From OKRs

So there you are: four key elements to take into account before launching your OKR implementation program.

In the next article in this series, we are going to look at how to build good OKRs. See you there!